Information push method and apparatus

Abstract

The present invention is applicable to the field of information processing technologies, and provides an information push method and apparatus. The method includes: acquiring historical behavior information of a user from a social data source; dividing, according to a preset rule, the acquired historical behavior information into one or more documents related to user behavior information; obtaining a model according to the document and by using a statistical learning method; and generating push information based on the model, and sending the push information to a client where a corresponding user is located. The push information in the present invention is generated based on the historical behavior information of the user, so that accuracy of information push can be effectively improved.

What is claimed is: 
     
       1. An information push method comprising:
 acquiring historical behavior information of a user from a social data source; 
 dividing, according to a preset rule, the acquired historical behavior information into one or more documents related to user behavior information; 
 establishing an initial model according to the one or more documents; 
 extracting characteristic information of each document using a statistical learning method; 
 obtaining a joint probability of the documents according to the extracted characteristic information; 
 maximizing the joint probability to obtain a parameter of the initial model; 
 obtaining a final model according to the parameter of the initial model; and 
 generating push information based on the final model, and sending the push information to a client where the user is located. 
 
     
     
       2. The method according to  claim 1 , wherein the dividing, according to the preset rule, the acquired historical behavior information into the one or more documents related to the user behavior information further comprises:
 dividing, according to a user identifier and time, the acquired historical behavior information into the one or more documents related to the user behavior information. 
 
     
     
       3. The method according to  claim 1 , wherein the extracting the characteristic information further comprises:
 extracting the characteristic information of each document according to at least one of (a) content, and (b) author information of each document. 
 
     
     
       4. The method according to  claim 1 , wherein before the generating the push information based on the final model, the method further comprises:
 detecting current behavior information of the user, wherein 
 correspondingly, the generating the push information based on the final model comprises: 
 generating, based on the final model and the detected current behavior information of the user, the push information matching the current behavior information of the user. 
 
     
     
       5. The method according to  claim 4 , wherein the generating, based on the final model and the detected current behavior information of the user, the push information matching the current behavior information of the user, and sending the push information to the client where the user is located further comprises:
 generating a list of possible user behaviors based on the final model and the detected current behavior information of the user; 
 scoring a possible user behavior in the list of possible user behaviors; and 
 pushing first N possible user behaviors with a highest score to the client where the user is located, wherein the N is greater than or equal to 1. 
 
     
     
       6. The method according to  claim 1 , wherein the historical behavior information of the user comprising at least one of: information browsing, information reposting, original information publishing, and information commenting. 
     
     
       7. The method according to  claim 1 , wherein each of the documents refers to content published by the user. 
     
     
       8. The method according to  claim 1 , wherein the social data source comprising the basic information of the user, the historical behavior information of the user, and friend information of the user. 
     
     
       9. An information push apparatus, comprising a processor and a non-transitory processor-readable medium having processor-executable instructions stored thereon, when executed by the processor, cause the processor to carry out the following operations:
 acquiring historical behavior information of a user from a social data source; 
 dividing, according to a preset rule, the historical behavior information acquired by the behavior information acquiring unit into one or more documents related to user behavior information; 
 establishing an initial model according to the one or more documents and extract characteristic information of each document by using a statistical learning method, obtain a joint probability of the documents in the one or more documents according to the extracted characteristic information, maximize the joint probability to obtain a parameter of the initial model, obtain a final model according to the parameter of the initial model; and 
 generating push information based on the final model, and send the push information to a client where the user is located. 
 
     
     
       10. The apparatus according to  claim 9 , wherein the processor is further configured to:
 divide, according to a user identifier and time, the acquired historical behavior information into the one or more documents related to the user behavior information. 
 
     
     
       11. The apparatus according to  claim 9 , wherein the processor is further configured to:
 extract the characteristic information of each document according to at least one of (a) content and (b) author information of each document. 
 
     
     
       12. The apparatus according to  claim 9 , wherein the processor is further configured to:
 detect current behavior information of the user, wherein 
 generate, based on the final model and the detected current behavior information of the user, the push information matching the current behavior information of the user, and send the push information to the client where the user is located. 
 
     
     
       13. The apparatus according to  claim 12 , wherein the processor is further configured to:
 generate a list of possible user behaviors based on the final model and the detected current behavior information of the user; 
 score a possible user behavior in the list of possible user behaviors; and 
 push first N possible user behaviors with a highest score to the client where the user is located, wherein the N is greater than or equal to 1. 
 
     
     
       14. The apparatus according to  claim 9 , wherein the historical behavior information of the user comprising at least one of: information browsing, information reposting, original information publishing, and information commenting. 
     
     
       15. The apparatus according to  claim 9 , wherein each of the documents refers to content published by the user. 
     
     
       16. An information push system, comprising:
 a crawler server, a parsing server, an information push server, and a client; 
 the crawler server, the parsing server, the information push server, and the client are connected to and communicate with each other in a wired or wireless manner; 
 the crawler server is configured to collect basic information of a user, historical behavior information of the user, from a social networking site, and send the collected basic information and historical behavior information to the parsing server; 
 the parsing server is configured to parse the basic information and the historical behavior information of the user into structured data, and send the structured data obtained to the information push server 
 the information push server is configured to divide the structured data into one or more sessions according to a preset rule, wherein each session comprising one or more documents related to a piece of behavior information of the user; establish an initial model according to the one or more sessions and by using a statistical learning method, extract characteristic information of each document, calculate a joint probability of the documents in the one or more sessions according to the extracted characteristic information, maximize the joint probability to obtain a parameter of the initial model, obtain a final model according to the parameter of the initial model, generate push information based on the obtained final model, and send the push information to the client. 
 
     
     
       17. The information push system according to  claim 16 , wherein the crawler server, the parsing server, and the information push server are server end formed by several functional servers together.